MOAK: The Mother of All Karlsonators (aka The Magnificent Monster)

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Founder of XSA-Labs
Joined 2012
Paid Member
The "Karlson" design was discredited many years ago (see Karlson Enclosure - List Thread) for a discussion that dates back to 1998), so its curious to see this type of design still popping up from time to time. Characteristics seem to be rough performance in its pass-band, and the "punch" is likely due to an undersized enclosure and/or a high Fb.

Discredited a design that achieves an efficient and high SPL (95 to 98 dB at 2.83V) over a huge bandwidth (55 Hz to 2.5 kHz)?! Look at the "Speaker that Kicks Butt in Large Spaces" thread for more info.
http://www.diyaudio.com/forums/full-range/237948-speaker-kicks-butt-large-spaces.html

Anyhow, this is the Karlsonator - not Karlson K15 which is a totally different animal. The Karlson curved aperture is used to reduce the effects of resonance peaks due to a hard edge at the terminus.
 
Founder of XSA-Labs
Joined 2012
Paid Member
2.5x Scaled Karlsonator with dual AE TDM15H's

I know you have been waiting for this... the MOAK version 1 with dual TDM15H's does a respectable job as a sub.

I scaled the height and depth by 2.5x, and the width by 1.5x, and enlarged the vent connecting the front and rear chamber by 2.4x.

Here is SPL vs Freq for 2.83 volts:

413879d1398086353-karlsonator-subwoofer-aka-magnificent-monster-2.5h_1.5w_2.4vent-karlsonator-dual-aetdm15h-freq-1m.png


Here is the Impedance:

413880d1398086353-karlsonator-subwoofer-aka-magnificent-monster-2.5h_1.5w_2.4vent-karlsonator-dual-aetdm15h-impedance.png


Here is the displacement:

413881d1398086353-karlsonator-subwoofer-aka-magnificent-monster-50inx15in-karlsonator-aetd15h-displ.png


The 14 mm xmax is reached at 42 volts (440 watts into 4 ohm load), the corresponding SPL vs Freq is here:

413882d1398086353-karlsonator-subwoofer-aka-magnificent-monster-2.5h_1.5w_2.4vent-karlsonator-dual-aetdm15h-freq-1m-xmax-42volts.png


The bass extension (-3 dB point is 20.5 Hz) - not too bad... :D the efficiency could be better as this is not an optimized design by any means.

Interestingly, playing with the Karlson aperture and making it go to a zero width cusp vs a finite width, reduces the peaks but efficiency is somewhat reduced:

413883d1398086692-karlsonator-subwoofer-aka-magnificent-monster-2.5h_1.5w_2.4vent-karlsonator-dual-aetdm15h-freq-1m-xmax-42volts-zero-point-slit.png


If you apply a -24 dB/oct Butterworth high pass at 10 Hz and a -48d dB/oct Butterworth low pass filter at 100 Hz, here is what you get:

413885d1398086992-karlsonator-subwoofer-aka-magnificent-monster-2.5h_1.5w_2.4vent-karlsonator-dual-aetdm15h-freq-1m-xmax-42volts-zero-point-slit-hpf-lpf.png


If you were to run this speaker in fullrange, here is the impulse response it is capable of producing - about as good as a sealed box with a 0.47 ms delay:

413887d1398087424-karlsonator-subwoofer-aka-magnificent-monster-2.5h_1.5w_2.4vent-karlsonator-dual-aetdm15h-impulse-xmax-42volts-zero-point-slit.png


So, as a starting point, the Karlsonator alignment is extremely versatile and can be scaled up or down like nobody's business. I think GregB will get a kick out of this :)
 

Attachments

  • 2.5h_1.5w_2.4vent-Karlsonator-Dual-AETDM15H-Freq-1m.png
    2.5h_1.5w_2.4vent-Karlsonator-Dual-AETDM15H-Freq-1m.png
    27.4 KB · Views: 852
  • 2.5h_1.5w_2.4vent-Karlsonator-Dual-AETDM15H-Impedance.png
    2.5h_1.5w_2.4vent-Karlsonator-Dual-AETDM15H-Impedance.png
    26.3 KB · Views: 834
  • 50inx15in-Karlsonator-AETD15H-Displ.png
    50inx15in-Karlsonator-AETD15H-Displ.png
    28 KB · Views: 838
  • 2.5h_1.5w_2.4vent-Karlsonator-Dual-AETDM15H-Freq-1m-at-xmax-42volts.png
    2.5h_1.5w_2.4vent-Karlsonator-Dual-AETDM15H-Freq-1m-at-xmax-42volts.png
    27.1 KB · Views: 838
  • 2.5h_1.5w_2.4vent-Karlsonator-Dual-AETDM15H-Freq-1m-at-xmax-42volts-zero-point-slit.png
    2.5h_1.5w_2.4vent-Karlsonator-Dual-AETDM15H-Freq-1m-at-xmax-42volts-zero-point-slit.png
    26.8 KB · Views: 1,052
  • 2.5h_1.5w_2.4vent-Karlsonator-Dual-AETDM15H-Freq-1m-at-xmax-42volts-zero-point-slit-HPF-LPF.png
    2.5h_1.5w_2.4vent-Karlsonator-Dual-AETDM15H-Freq-1m-at-xmax-42volts-zero-point-slit-HPF-LPF.png
    27.9 KB · Views: 2,487
  • 2.5h_1.5w_2.4vent-Karlsonator-Dual-AETDM15H-Impulse-at-xmax-42volts-zero-point-slit.png
    2.5h_1.5w_2.4vent-Karlsonator-Dual-AETDM15H-Impulse-at-xmax-42volts-zero-point-slit.png
    18.8 KB · Views: 820
Last edited:
Well, at least in my simulations, the aperture has a noticeable effect if you look at the plot showing how the peaks were smoothed out with an aperture that goes to a sharp cusp vs a slit.

You like this?

An externally hosted image should be here but it was not working when we last tested it.


An externally hosted image should be here but it was not working when we last tested it.
 
Founder of XSA-Labs
Joined 2012
Paid Member
What application are you using to perform those simulations?

It is a script that I wrote in Akabak - details of how I modeled the K15 and the Karlson aperture are here: http://www.diyaudio.com/forums/full-range/213594-karlson-10.html#post3528706

The Karlsonator uses the same block of code for modeling the front aperture in the K15 model - a series of 9 rectangular vents with a profile prescribed by the Karlson aperture curve function. The rest of the Karlsonator is a series of waveguides - like a folded tapped horn.

There was a study I did leading up to my model of the K aperture where I looked at drilling strategically placed holes in a TL - like a flute to control response (http://www.diyaudio.com/forums/full-range/237384-flute-pipe-tl.html). That model gave me the methodology to apply to K aperture. It is quite a complex piece of code - not trivial by any means and only possible with Akabak, Akabak 3d, or a FEM solver like COMSOL Multiphysics.

The speaker electrical impedance plots that I have simulated closely match the shape and locations of measurements provided by Freddi. I have gone down this path of people questioning the model before in the other thread - like asking what "knobs" I have to tweak to get the response? The thinking is a model can do anything given enough knobs to adjust. there are *NO* knobs other than physical dimensions and amount of damping - which does not alter things that much other than to smooth peaks.
 
Last edited:
Hi xrk971,

Could you, please, simulate the output of the same enclosure, but w/ something like Art's (weltersys) Keystone exit. Art did a lot of measurements to compare all kinds of different exit shapes (including the Karlson slot), and ended up w/ the Keystone shape providing an advantage. Maybe less opening height, a much bigger slot top, and a reduced from internal width bottom of the Keystone, arriving at a similar terminus area.

http://www.diyaudio.com/forums/subwoofers/185588-keystone-sub-using-18-15-12-inch-speakers.html

Regards,
 
the slit portion of the aperture is important - if I block a few inches of the top of this regular type K18's aperture, subjective highs, "speed" and system damping are affected. btw, a typical 2-way Karlson should employ a K-tube tweeter. The "X15" Karlson 2-way system 1965 introduced the slotted treble waveguide.

I learned over a decade ago htat the original Karlson K15 of 1951 had virtues and subjectively sounded better than a lot of mediocre Harmon products:)

An externally hosted image should be here but it was not working when we last tested it.


hak6.5 build by AmadeusMozart - the design is based on one of Carl's builds
XK6.5front.jpg

XDy0Lk3.jpg

one of Carl's test boxes loaded with an 8
An externally hosted image should be here but it was not working when we last tested it.

1965 X15 ad sheet
An externally hosted image should be here but it was not working when we last tested it.
 
Last edited:
It is a script that I wrote in Akabak - details of how I modeled the K15 and the Karlson aperture are here: http://www.diyaudio.com/forums/full-range/213594-karlson-10.html#post3528706

The Karlsonator uses the same block of code for modeling the front aperture in the K15 model - a series of 9 rectangular vents with a profile prescribed by the Karlson aperture curve function. The rest of the Karlsonator is a series of waveguides - like a folded tapped horn.

Thanks, I will take a closer look at that thread.

What sticks out at me immediately with these things is the decidedly non-flat response in the passband. Why go through all that trouble involved in a complex build to get a result like that? If I wanted to use it to "produce" music, perhaps it's something I'd consider, but it's not something I'd look at if I wanted to "reproduce" music. This is probably why I never considered the Karlson aperture more than a curiosity

I suspect that it's possible to get the response you're looking for with the same driver is a much more "traditional" alignment.

The "theory" of the Karlson aperture can be seen here: The Karlson Homepage. The firs thing I noticed was no impedance or acoustic measurements to back up the claim of "broadband resonance" or "widening peaks".

I'll read a bit further. Thanks for the link.
 
Hi freddi,

Nice pictures, but:

I don't have anything for or against the Karlson type enclosure, but the promo you just posted says: "...true-pitch bass down to 12cps...High Efficiency (approx. 40%)...Only about 1/8" cone travel required for highest levels..." Come on :).

I also question it's applicability for true subwoofer applications (but a lot of designs posted here under subwoofers are just woofers, or supplemental woofers).

Regards,
 
Could you, please, simulate the output of the same enclosure, but w/ something like Art's (weltersys) Keystone exit. Art did a lot of measurements to compare all kinds of different exit shapes (including the Karlson slot), and ended up w/ the Keystone shape providing an advantage. Maybe less opening height, a much bigger slot top, and a reduced from internal width bottom of the Keystone, arriving at a similar terminus area.
Oliver,

xrk971's designs have a higher aspect ratio than the Keystone tapped horn, so the more narrow top "Karlson" shaped exit may work better than it did in the "stubbier" TH.

xrk971's simulations of 9 rectangular vents could approximate any of the exit apertures I tested, if he is interested in seeing how the simulations compare with measured results on any of the shapes tested below, I can provide the measurements for any of "A" to "V". I tested "in between" versions of most pictured as well, but as the differences were minor, did not save those results, as I would have run out of alphabet ;).

The ground plane reflection of a high aspect ratio (tall, skinny) box with these type of exits influences response, the measured response is quite different when the narrow portion of the aperture is placed down (near the ground plane) rather than up. An Akaback script won't "reflect" those differences, unless the mouth orientation in relation to the ground plane can be specified.

Back to Deon, I would also advise against a vertical driver arrangement of more than two drivers in a TH alignment using a "Karlson" shaped exit (especially a high aspect ratio "V" shape) for the same reasons GM mentioned in #4 and #7. Even in a BR alignment, I'd advise against it, though the pressure differentials would be far less.

Art
 

Attachments

  • Exit Shapes A-V.jpg
    Exit Shapes A-V.jpg
    238.1 KB · Views: 234
Last edited:
Hi Art,

Post #35: "...The ground plane reflection of a high aspect ratio (tall, skinny) box with these type of exits influences response, the measured response is quite different when the narrow portion of the aperture is placed down (near the ground plane) rather than up. An Akaback script won't "reflect" those differences, unless the mouth orientation in relation to the ground plane can be specified..."

From what I understand xrk971 takes all those things into consideration, and more. I wish I could take a look at the whole script, but it is obviously proprietary, and I might just get more confused about AkAbak than I already am. :) It's good to know that you have all those measurements, maybe xrk971 will pick up on that.

Regards,
 
Founder of XSA-Labs
Joined 2012
Paid Member
An Akaback script won't "reflect" those differences, unless the mouth orientation in relation to the ground plane can be specified.

I can specify the orientation and turn on ground reflections which has a big effect. For initial design development I turn reflections off so that the speaker's inherent output in an "anechoic" chamber can be seen. From there optimize the speaker placement and sometimes mouth orientation to achieve a flat output with reflections turned on. For example, here is the MOAK v1 with reflections turned on with rear wall 60 in away from back of speaker, and speaker sitting on floor with center point between both drivers at 16 inches above floor:

413925d1398099838-karlsonator-subwoofer-aka-magnificent-monster-2.5h_1.5w_2.4vent-karlsonator-dual-aetdm15h-freq-1m-xmax-42volts-zero-point-slit-refelctions-.png


Note the bass enhancement from the floor and the ripples caused by the reflections interfering with main output.
 

Attachments

  • 2.5h_1.5w_2.4vent-Karlsonator-Dual-AETDM15H-Freq-1m-at-xmax-42volts-zero-point-slit-Refelctions-.png
    2.5h_1.5w_2.4vent-Karlsonator-Dual-AETDM15H-Freq-1m-at-xmax-42volts-zero-point-slit-Refelctions-.png
    27.4 KB · Views: 1,532
Last edited:
Founder of XSA-Labs
Joined 2012
Paid Member
What sticks out at me immediately with these things is the decidedly non-flat response in the passband. Why go through all that trouble involved in a complex build to get a result like that?

Non-flat because you see ripples and dips and peaks that are +/- 3dB - a very good performer for many speakers and especially given the broad bandwidth. Here is what this speaker can do in a wide band application with the addition of a 6 ohm + 10 mH BSC circuit (or EQ'ing):

413926d1398100304-karlsonator-subwoofer-aka-magnificent-monster-2.5h_1.5w_2.4vent-karlsonator-dual-aetdm15h-freq-1m-xmax-42volts-zero-point-slit-refelctions-.png


Can you find another speaker with high SPL capability that has -3dB point of 18 Hz on the low end and goes up to 1.16 kHz on the high end? That is 6 octaves!

Add a horn loaded CD tweeter and you have a nice 2-way that can rock the house (or stadium).

And here is the corresponding Impulse Response for this setup:

413935d1398100854-karlsonator-subwoofer-aka-magnificent-monster-2.5h_1.5w_2.4vent-karlsonator-dual-aetdm15h-ir-xmax-42volts-zero-point-slit-refelctions-60in-.png
 

Attachments

  • 2.5h_1.5w_2.4vent-Karlsonator-Dual-AETDM15H-Freq-1m-at-xmax-42volts-zero-point-slit-Refelctions-.png
    2.5h_1.5w_2.4vent-Karlsonator-Dual-AETDM15H-Freq-1m-at-xmax-42volts-zero-point-slit-Refelctions-.png
    27.3 KB · Views: 803
  • 2.5h_1.5w_2.4vent-Karlsonator-Dual-AETDM15H-IR-at-xmax-42volts-zero-point-slit-Refelctions-60in-.png
    2.5h_1.5w_2.4vent-Karlsonator-Dual-AETDM15H-IR-at-xmax-42volts-zero-point-slit-Refelctions-60in-.png
    17.8 KB · Views: 797
Last edited:
Can you find another speaker with high SPL capability that has -3dB point of 18 Hz on the low end and goes up to 1.16 kHz on the high end?

What does the polar response of that look like?

In any case, given that wide bandwidth claim, I want to see the actual FR measurements to back that up, not a sim. As I indicated before, all of the Karlson FR curves I've seen show a decidedly non-flat response, and many sims fall apart above 200 Hz or so because of how the driver itself is modeled in the sim.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.